High-Resolution Multimodal Reasoning on HR-Bench 8K FSP
94.8ACCRTWI
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| RTWIModel Variant=Qwen3-VL Instruct, Online setting=true2026.02 | 94.8 | 60.3 | |
| DeepconfModel Variant=Qwen3-VL Instruct, Online setting=true2026.02 | 93.5 | 58.9 | |
| SCModel Variant=Qwen3-VL Instruct, Online setting=true2026.02 | 93 | — | |
| ASCModel Variant=Qwen3-VL Instruct, Online setting=true2026.02 | 93 | 54.6 | |
| ESCModel Variant=Qwen3-VL Instruct, Online setting=true2026.02 | 93 | 51.7 | |
| CISCModel Variant=Qwen3-VL Instruct, Online setting=true2026.02 | 91 | 53.6 | |
| BaseModel Variant=Qwen3-VL Instruct, Online setting=true2026.02 | 90.8 | — | |
| RTWIModel Variant=Qwen3-VL Thinking, Online setting=true2026.02 | 88.8 | 59.3 | |
| Self-Cer.Model Variant=Qwen3-VL Instruct, Online setting=true2026.02 | 88.8 | 57.1 | |
| Self-Cer.Model Variant=Qwen3-VL Thinking, Online setting=true2026.02 | 88 | 43.8 | |
| CISCModel Variant=Qwen3-VL Thinking, Online setting=true2026.02 | 87.8 | 42.6 | |
| DeepconfModel Variant=Qwen3-VL Thinking, Online setting=true2026.02 | 87 | 46.9 | |
| ThymeOnline setting=true2026.02 | 86.5 | — | |
| DeepEyesOnline setting=true2026.02 | 86.5 | — | |
| SCModel Variant=Qwen3-VL Thinking, Online setting=true2026.02 | 86.5 | — | |
| ASCModel Variant=Qwen3-VL Thinking, Online setting=true2026.02 | 86.5 | 50.4 | |
| ESCModel Variant=Qwen3-VL Thinking, Online setting=true2026.02 | 86.5 | 35.7 | |
| BaseModel Variant=Qwen3-VL Thinking, Online setting=true2026.02 | 78.3 | — | |
| GPT-4oOnline setting=true2026.02 | 60.8 | — |